Related Experiment Video
Updated: May 26, 2026
11:19
Pairwise Growth Competition Assay for Determining the Replication Fitness of Human Immunodeficiency Viruses
Published on: May 4, 2015
Developing a predictive model to prioritize human immunodeficiency virus partner notification in North Carolina
Brooke E Hoots1, Pia D M MacDonald, Lisa B Hightow-Weidman
1Department of Epidemiology, Gillings School of Global Public Health, University of North Carolina, Chapel Hill, NC, USA.
Sexually Transmitted Diseases
|December 21, 2011
Summary
A new model helps disease intervention specialists prioritize human immunodeficiency virus (HIV) testing by predicting undiagnosed infections in sexual partners, improving resource allocation for partner notification.
Area of Science:
- Epidemiology
- Public Health
- Biostatistics
Background:
- Disease intervention specialists (DIS) face time constraints and increased workloads due to rising human immunodeficiency virus (HIV) testing.
- Prioritizing partner notification is challenging with competing responsibilities.
Purpose of the Study:
- To develop a predictive model for identifying undiagnosed HIV infections in sexual partners.
- To optimize the allocation of resources for DIS interviews.
Main Methods:
- Utilized demographic, behavioral, and partnership data from DIS records in North Carolina (2003-2007).
- Employed multiple logistic regression with generalized estimating equations to create a predictive model and risk scores.
- Assessed algorithm performance using sensitivities and specificities at various risk score cutoffs.
Main Results:
- Identified five key factors predicting partnerships with undiagnosed HIV: short diagnosis-to-interview time, no crack use history, no anonymous sex, fewer total partners, and age difference between partners.
- The model allows DIS to select appropriate cutoffs for partner location based on the balance of false negatives and false positives.
Conclusions:
- The model, despite low overall predictive power, can significantly reduce the number of partners needing interviews while maintaining high sensitivity.
- It serves as a valuable tool for identifying partners requiring more intensive resource allocation for locating.
